Confusion Isn't Everything

Transcription

1 Scholarship Repository University of Minnesota Law School Articles Faculty Scholarship 2013 Confusion Isn't Everything William McGeveran University of Minnesota Law School, Mark P. McKenna Follow this and additional works at: Part of the Law Commons Recommended Citation William McGeveran and Mark P. McKenna, Confusion Isn't Everything, 89 Notre Dame L. Rev. 253 (2013), available at This Article is brought to you for free and open access by the University of Minnesota Law School. It has been accepted for inclusion in the Faculty Scholarship collection by an authorized administrator of the Scholarship Repository. For more information, please contact

2 CONFUSION ISN T EVERYTHING William McGeveran* and Mark P. McKenna** The typical shorthand justification for trademark rights centers on avoiding consumer confusion. But in truth, this encapsulation mistakes a method for a purpose: confusion merely serves as an indicator of the underlying problems that trademark law seeks to prevent. Other areas of law accept confusion or mistake of all kinds, intervening only when those errors lead to more serious harms. Likewise, every theory of trademark rights considers confusion troubling solely because it threatens more fundamental values such as fair competition or informative communication. In other words, when it comes to the deep purposes of trademark law, confusion isn t everything. Yet trademark law s structure now encourages courts to act otherwise, as if confusion itself were the ultimate evil with which trademark law is concerned and as if its optimal level were zero. Trademark adjudication increasingly fetishized confusion over the last half century while simultaneously expanding its scope to cover dramatically more situations. As a result, trademark rights conflicted more frequently with open markets and free speech. Even when special doctrines developed to protect those values, they tended to devolve back into confusion analysis. In many situations often including examples such as trade dress, comparative advertising, or uses in expressive works competing values get short shrift from a framework preoccupied with eradicating confusion of any description. This Article analyzes causes of this phenomenon, problems that result, and possible cures. It begins with a historical analysis demonstrating how the growth of confusion reasoning overwhelmed inherent limits on trademark rights that once protected competition or communication values. It then explains the resulting damage. Sometimes courts penalize socially valuable but 2013 William McGeveran & Mark P. McKenna. Individuals and nonprofit institutions may reproduce and distribute copies of this Article in any format at or below cost, for educational purposes, so long as each copy identifies the author, provides a citation to the Notre Dame Law Review, and includes this provision in the copyright notice. We received many helpful comments from numerous colleagues as we developed the ideas in this Article. We are particularly grateful to Tom Cotter, Deven Desai, Dan Gifford, Eric Goldman, Laura Heymann, Mark Lemley, Ruth Okediji, Lisa Ramsey, Jay Tidmarsh, and Rebecca Tushnet for their thoughts. Thanks also to participants in conferences and workshops where one or both of us presented earlier versions of this project, including the 2011 Intellectual Property Scholars Conference, the 2012 Works-in- Progress in Intellectual Property conference, the academic track at the 2012 International Trademark Association Annual Meeting, the Berkeley IP Theory Seminar, the 2012 UCLA Entertainment, Media & IP Colloquium, and workshops at the Notre Dame Law School, the University of Minnesota Law School, and the Vanderbilt Law School. Finally, we appreciate research assistance from Catherine Mitros and Vincent Pham. * Associate Professor and Vance Opperman Research Scholar, University of Minnesota Law School. ** Associate Dean for Faculty Development, Professor of Law, and Notre Dame Presidential Fellow, Notre Dame Law School. 253

3 254 notre dame law review [vol. 89:1 unlicensed uses of marks; even when they probably would not find the uses infringing, the high administrative costs imposed on defendants discourage such uses. Finally, the Article proposes a better alternative. Courts should identify categories of cases implicating competition or communication values at the outset of litigation and should handle them differently than ordinary cases. Sometimes courts should simply ignore confusion when enforcing trademark rights would cause too much damage to other values. In other cases, they should distinguish between types of confusion; when balanced against important competing values, minimal or temporary confusion need not be given the same weight as outright deception that diverts trade. Our recommended approach offers stronger and earlier protection for competition and communication. It also puts confusion detection back in its proper place not as the ultimate purpose of trademark law, but its tool. INTRODUCTION Ask any law student to describe the purpose of trademark law and you re likely to get a simple answer: preventing consumer confusion. Indeed, you d likely get the same simple answer from most trademark lawyers. Following in this path, courts tend to view confusion itself as the ill that trademark law seeks to cure and to assume that the optimal level of confusion is always zero. Both these propositions are wrong. The resulting errors undermine important values, such as fair competition and effective communication, which should lie at the heart of trademark law s protection. The first proposition mistakes trademark law s mediate purpose for its ultimate purpose. Consumer confusion is but a symptom of the disease targeted by trademark law. Confusion ought to serve as a heuristic device to identify situations where certain bad consequences may follow, including diversion of trade and distorted purchasing decisions. 1 A small amount of confusion does not inevitably cause problems, and it might even prove beneficial in some cases where it serves as a prelude to greater understanding (which, as we repeatedly tell our first-year law students, it often does). Increasingly, however, courts have come to see elimination of consumer confusion itself as the alpha and omega of trademark law. This trend has increased ever since courts began developing multi-factor tests for likelihood of confusion to measure infringement. 2 The Lanham Act now expressly identifies confusion as a basis for liability. 3 And the scope of that confusion, and therefore of trademark rights, has expanded radically. Courts (and later Congress) steadily added to the types of actionable confusion: concerning more relationships (such as sponsorship or affiliation ); in connection with more protectable subject matter (such as product packaging and design); and beyond the point of sale (such as initial interest 1 See Mark A. Lemley & Mark McKenna, Irrelevant Confusion, 62 STAN. L. REV. 413 (2010). 2 For a history of the development of the multi-factor likelihood of confusion test, see Robert G. Bone, Taking the Confusion Out of Likelihood of Confusion : Toward a More Sensible Approach to Trademark Infringement, 106 NW. U. L. REV (2012) U.S.C. 1114, 1125(a) (2006).

4 2013] confusion isn t everything 255 or post-sale confusion). 4 This inexorable growth brought trademarks into much more frequent conflict with previously acceptable uses often uses that serve other crucial values of a well-functioning trademark system. The singular focus on confusion eventually displaced previous understandings of a trademark s purpose. 5 By 2013, experienced judges on the Second Circuit thought they were underscoring the absurdity of a fair use test applied by the district court when they pointed out that it would lead to the dismissal of these claims without addressing what is beyond doubt the central question in considering consumer confusion: whether consumers were actually confused by the allegedly infringing product. 6 Not only is this holding tautological, it assumes that every trademark case must turn on confusion. When it comes to the deep purposes of trademark law, however, confusion isn t everything. The second error noted above that the law always should strive to eliminate all confusion entirely is related to the first. Courts routinely interpret statutory requirements by considering the classic question of the mischief sought to be avoided by the law. 7 This is especially so in an open-ended statute like the Lanham Act that delegates interpretive authority to the courts. 8 But if the mischief to be avoided by trademark law is confusion itself, courts understandably reason, then surely they should stamp it out whenever it appears. 9 This habit persists despite contrary indications in trademark doctrine. In a case involving descriptive fair use, the Supreme 4 For a full account of the expansion of confusion, see infra Section I.B. 5 See Mark P. McKenna, The Normative Foundations of Trademark Law, 82 NOTRE DAME L. REV (2007). 6 Kelly-Brown v. Winfrey, 717 F.3d 295, 307 (2d Cir. 2013) AM. JUR. 2D Statutes 71 (2012) ( In the construction of an ambiguous statute, it is proper to take into consideration the particular evils at which the legislation is aimed, or the mischief sought to be avoided.... (footnote omitted)); see Awotin v. Atlas Exch. Nat l Bank of Chi., 295 U.S. 209, 214 (1935) ( The prohibition would be nullified and the evil sought to be avoided would persist.... Such a construction of the statute is inadmissible. ); Holy Trinity Church v. United States, 143 U.S. 457, 463 (1892) ( [A]nother guide to the meaning of a statute is found in the evil which it is designed to remedy.... ). 8 Pierre N. Leval, Trademark: Champion of Free Speech, 27 COLUM. J.L. & ARTS 187, 198 (2004). 9 While there is no absolute quantitative threshold for determining what level of confusion is actionable, courts have been persuaded by surveys showing 15% confusion, and sometimes even less. See, e.g., Exxon Corp. v. Tex. Motor Exch. of Hous., Inc., 628 F.2d 500, 507 (5th Cir. 1980) (finding a 15% level of confusion strong evidence of likelihood of confusion); RJR Foods, Inc. v. White Rock Corp., 603 F.2d 1058, 1061 (2d Cir. 1979) (upholding the district court s finding that a 15 20% level of confusion corroborates likelihood of confusion); James Burrough Ltd. v. Sign of the Beefeater, Inc., 540 F.2d 266, 279 (7th Cir. 1976) (finding that a 15% level of confusion was neither small nor de minimis). In one case, the court called evidence of 8.5% confusion strong evidence of a likelihood of confusion. Grotrian, Helfferich, Schulz, Th. Steinweg Nachf. v. Steinway & Sons, 365 F. Supp. 707, 716 (S.D.N.Y. 1973), modified on other grounds, 523 F.2d 1331 (2d Cir. 1975). Lest those cases appear isolated and old, this past summer the court in Gucci America, Inc. v. Guess?, Inc., 868 F. Supp. 2d 207, 232, 248 (S.D.N.Y. 2012), held that the defendant s products infringed because they were likely to cause post-sale confusion after finding that

5 256 notre dame law review [vol. 89:1 Court explicitly held that some possibility of consumer confusion must be compatible with fair use. 10 Functionality doctrine similarly disregards confusion when its requirements are otherwise satisfied. 11 This wisdom is not woven into the fabric of trademark law very well, however. Almost every so-called defense in trademark law relies in the end on the absence of confusion rather than the presence of important alternative values. 12 And the expansive definition of confusion means markholders often can claim some malign influence on consumers. We suggest historical reasons for this doctrinal circularity below. 13 Whatever the explanation, however, contemporary courts lack tools to consider values other than confusion because they understand the elimination of all confusion to be the task at hand. This Article tackles both of these related misunderstandings of consumer confusion in trademark law: the supposed role of confusion as the law s sole driving purpose and the imperative to reduce it to zero. We trace the development of the preoccupation with confusion, demonstrate its negative consequences, and suggest avenues for reform. We begin with history. Why do so many courts and other actors see elimination of consumer confusion as the primary (perhaps even the sole) mission of trademark law? Part I describes how courts traditionally relied on the inherent limits of more narrowly defined trademark rights to protect socially beneficial uses of a mark, obviating the need for special doctrines. We then show how those inherent limits eroded over time with the expansion of rights. Trademark adjudication increasingly fetishized confusion over the last half century while simultaneously expanding the scope of that concept to cover dramatically more situations. These changes created more conflicts between confusion and other values, yet left trademark law without the capacity to adjudicate those conflicts efficiently. We then elaborate on the problems caused by this state of affairs. Part II identifies other values, particularly fair competition and open communication, which can be threatened by trademark structures overly reliant on eradicating confusion. It also explains how courts attempts to articulate new limits on the scope of trademark rights largely failed because they recursively relied on assessments of the now-overgrown confusion concept. Cases involving socially valuable, but unlicensed, uses of marks present two distinct problems. First, current doctrine submerges the other values and perpetuates the notion that preventing consumer confusion is itself the aim of trademark law. In response, courts tend systematically to overweigh confusion, the surveys showed a maximum level of net confusion among casual observers in the postsale setting of only 5.8%. 10 KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 543 U.S. 111, 121 (2004). 11 See TrafFix Devices, Inc. v. Mktg. Displays, Inc., 532 U.S. 23, (2001); Eppendorf-Netheler-Hinz GMBH v. Ritter GMBH, 289 F.3d 351, 355 (5th Cir. 2002). 12 See William McGeveran, Rethinking Trademark Fair Use, 94 IOWA L. REV. 49, (2008). 13 See infra Part I.

6 2013] confusion isn t everything 257 which they see as more essential to trademark law than any of those other values. Second, and at least as important, courts cannot dispose of cases early in litigation because they generally cannot address competing values outside a fact-intensive likelihood of confusion analysis. As a result, even when an unlicensed use of a mark ultimately would be found lawful, the user faces the prospect of long and expensive litigation to validate that use. These costs discourage legitimate uses on the ground and increase the effectiveness of cease-and-desist letters, including those that claim stronger rights than the law really confers. Finally, in Part III we take up the task of reforming trademark adjudication to better recognize and protect other values. We propose a set of what we call limiting doctrines to do so. In civil procedure terms, some of these are affirmative defenses, but we also include other doctrinal structures that efficiently vindicate non-confusion interests served by a defendant s actions. 14 In brief summary, we suggest that courts should first identify categories of cases that raise competing values, and then should handle these identified types of cases differently at both the liability and remedy phases. The sorting would leave the handling of most trademark cases undisturbed, but would significantly change doctrine in the affected categories. At the liability phase, in some instances we propose categorical safe harbors for the defendant, while in others we recommend a much more limited role for confusion. At the remedy phase, courts should remember that eradication of all confusion is not their goal and limit their injunctive relief to those measures necessary, on balance, to prevent serious harm to consumers. Described at this level of generality, we realize that our premise can seem abstract. We will illustrate the approach with three fact patterns that recur frequently and present conflicts between the elimination of confusion on one hand, and interests in competition and communication on the other. First, the ability to offer certain product features can be essential to compete in a market, but use of those features may also give rise to trademark claims, particularly as protection extends to more varieties of trade dress. Functionality doctrine generally refuses protection for mechanical features like the dual-spring design of a mechanism for keeping outdoor signs upright despite adverse wind conditions. 15 It much less consistently preserves access to competitively important nonmechanical features such as the color on the sole of a fashionable shoe. 16 Second, unlicensed third parties often use a mark when they promote their own products to consumers. This situation arises most obviously in the form of comparative advertising, but it also can come 14 See William McGeveran, Life in the Fast Lane: Of Presumptions, Defenses, and Burdens, 1 IP THEORY 25, (2010), available at iss1/2/. So, for example, the requirement of nonfunctionality in the prima facie infringement case is a limiting doctrine in this sense. So, too, are some of our remedial options that are not full defenses from liability. See infra Part III. 15 See TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23, (2001). 16 See Christian Louboutin S.A. v. Yves Saint Laurent Am. Holding, Inc., 696 F.3d 206, 216 (2d Cir. 2012).

7 258 notre dame law review [vol. 89:1 up when sellers, repairers, or brokers use a mark to explain their own services to consumers.. 17 The classic fair use scenario, in which a third party uses a mark for its descriptive meaning in order to describe its own product or service, presents similar conflicts between competitive purposes and confusion reading. 18 Finally, unauthorized expressive uses of marks in media from oil paintings to movies and video games face trademark claims based on the possibility that their audiences will become confused about an association between the markholder and the expressive work. 19 We will return to these three product features, third-party promotions, and uses in expressive works as concrete examples of situations that limiting doctrines should protect more effectively. They embody core trademark values of fair competition (product features), open communication (uses in expressive works), or both (comparative advertising and other third-party promotions). As a final introductory note, we offer a quick preliminary response to a sensible question we have often received about this project: if the breadth of confusion is the root problem, why not simply make confusion narrower again? One of us has written extensively endorsing just such an approach, and both of us continue to think it would be wise to alter the confusion rationale along these lines. 20 Nevertheless, for several reasons we think limiting doctrines applied to particular uses also demand attention from courts, scholars, and reformers. First, more general efforts to scale back trademark rights may prove unsuccessful, and in any event reversing decades of trademark expansion would take time. Our proposal gives courts concrete tools they can use right now to protect other values when they are most seriously threatened by trademark enforcement. Second, current doctrine seldom calls upon courts to think hard about the differences between types of confusion or the varied types and degrees of harm they may cause, nor to consider other important values aside from confusion. Confronting courts more directly with the costs of trademark overprotection should help nudge them toward a more wide-ranging re-evaluation of confusion. Third, even with a more limited understanding of consumer confusion, there will always be times when other values ought to take precedence over confusion for the 17 See, e.g., Toyota Motor Sales, U.S.A., Inc. v. Tabari, 610 F.3d 1171, 1175 (9th Cir. 2010) (considering suit against automobile broker unaffiliated with Lexus who used <buya-lexus.com> as a domain name); August Storck K.G. v. Nabisco, Inc., 59 F.3d 616, (7th Cir. 1995) (upholding comparative use of competitor s mark on candy packaging). 18 See KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 543 U.S. 111, 114 (2004) (involving a dispute over variations of a single term found in one company s trademark that were used in a competitor s product description). 19 See, e.g., Univ. of Ala. Bd. of Trs. v. New Life Art, Inc., 683 F.3d 1266, 1269 (11th Cir. 2012) (alleging violation based on depiction of university football team logos and uniforms in oil paintings); E.S.S. Entm t 2000, Inc. v. Rock Star Videos, Inc., 547 F.3d 1095, 1097 (9th Cir. 2008) (claiming violation based on imitation of logo and building décor in video game). 20 See Lemley & McKenna, supra note 1, at ; Mark A. Lemley & Mark P. McKenna, Owning Mark(et)s, 109 MICH. L. REV. 137, (2010); see also McGeveran, supra note 12, at (criticizing focus on likelihood of confusion).

8 2013] confusion isn t everything 259 good of the entire trademark regime and the markets it supports. For example, courts already disregard even confusion about actual source the narrowest and most traditional form of confusion when they determine that the disputed feature was functional. 21 Given the unbroken history of expanding rights, limiting doctrines probably will be essential to protect other values important to the whole trademark system. That certainly doesn t mean we should ignore problems with the confusion rationale. But it does mean that we cannot afford to ignore limiting doctrines. I. HISTORICAL EVOLUTION At one time, the inherently limited nature of trademarks did much of the doctrinal work of trademark law. As the Supreme Court famously said almost a century ago in United Drug Co. v. Theodore Rectanus Co., [t]here is no such thing as property in a trade-mark except as a right appurtenant to an established business or trade in connection with which the mark is employed. 22 Put simply, trademark rights were naturally limited by their purposes; as a result, many uses of a mark by other parties were not even prima facie infringing. As this Part explains, courts traditionally relied on these natural limits to deny relief where the defendant s use did not implicate trademark law s concerns. That was generally enough to prevent serious conflicts between trademark rights and other values such as competition or speech. Over the last several decades, however, trademark rights have expanded so much that conflicts are now much more frequent, and the old doctrinal formulations are no longer sufficient to resolve them. A. The Erosion of Inherent Limits in Trademark Rights Until sometime in the early to mid-twentieth century, trademark law (along with the broader law of unfair competition of which it has always been a part) was designed for the limited purpose of preventing trade diversion by competitors. As Justice Strong wrote in Canal Co. v. Clark: [I]n all cases where rights to the exclusive use of a trade-mark are invaded, it is invariably held that the essence of the wrong consists in the sale of the goods of one manufacturer or vendor as those of another; and that it is only when this false representation is directly or indirectly made that the party who appeals to a court of equity can have relief. This is the doctrine of all the authorities J. THOMAS MCCARTHY, MCCARTHY ON TRADEMARKS AND UNFAIR COMPETITION 7:63 (4th ed. 2013) [hereinafter MCCARTHY] ( Functionality is a potent public policy, for it trumps all evidence of actual consumer identification of source and all evidence of actual consumer confusion caused by an imitator. ) U.S. 90, 97 (1918) U.S. 311, (1872); see also Lawrence Mfg. Co. v. Tenn. Mfg. Co., 138 U.S. 537, 546 (1891) (quoting the same passage from Canal Co. v. Clark). The essence of the wrong was the same in unfair competition cases that did not involve a technical trade-

9 260 notre dame law review [vol. 89:1 Nevertheless, courts in this era recognized that they had to be careful in denouncing trade diversion after all, capitalist competition consists inherently of efforts to divert others customers. Applied too broadly, unfair competition law could undermine the benefits of a market economy. Courts guarded against overextension by carefully distinguishing between legitimate forms of competition ( mere competition ) and illegitimate attempts to divert trade. 24 Deceptiveness was the point of demarcation: a defendant s use was actionable only if it deceived consumers and thereby diverted trade that otherwise would have gone to the plaintiff. 25 Because trademark rights were so narrow in this era, courts were more willing to presume deception in trademark infringement cases than in unfair competition, but both types of cases turned on deception that resulted in the passing off of the defendant s goods as those of the plaintiff. 26 mark. See Hanover Star Milling Co. v. Metcalf, 240 U.S. 403, 413 (1916) ( Th[e] essential element is the same in trade-mark cases as in cases of unfair competition unaccompanied with trade-mark infringement. ); JAMES LOVE HOPKINS, THE LAW OF TRADEMARKS, TRADENAMES AND UNFAIR COMPETITION 4, at 12 (3d ed. 1917) ( The principles involved in trademark cases and tradename cases have been substantially identical. ). 24 See, e.g., Coats v. Merrick Thread Co., 149 U.S. 562, 566 (1893) ( Rival manufacturers may lawfully compete for the patronage of the public in the quality and price of their goods, in the beauty and tastefulness of their enclosing packages, in the extent of their advertising, and in the employment of agents, but they have no right, by imitative devices, to beguile the public into buying their wares under the impression they are buying those of their rivals. ); Taylor v. Carpenter, 2 Sand. Ch. 603, 617 (N.Y. Ch. 1846) ( [Protection of trademarks] does not at all trench upon the rights of others, by a course of conduct equally deserving and praiseworthy, to enter the lists of competition, and bear off the palm. But it will not allow them by falsehood, fraud, and forgery, to filch from another his good name, and share it in common with him, or destroy or impair it. ). 25 See Borden Ice Cream Co. v. Borden s Condensed Milk Co., 201 F. 510, 513 (7th Cir. 1912) ( The deception of the public naturally tends to injure the proprietor of a business by diverting his customers and depriving him of sales which otherwise he might have made. This, rather than the protection of the public against imposition, is the sound and true basis for the private remedy. ); see also Hanover Star, 240 U.S. at ( This essential element is the same in trade-mark cases as in cases of unfair competition unaccompanied with trade-mark infringement. ); Crescent Tool Co. v. Kilborn & Bishop Co., 247 F. 299, 300 (2d Cir. 1917) (holding a second comer defendant liable where the public will believe his goods have come from the first [comer], and will buy, in part, at least, because of that deception ); Taylor v. Carpenter, 23 F. Cas. 742, 744 (C.C.D. Mass. 1844) (No. 13,785) ( The case presented is one of unmitigated and designed infringement of the rights of the plaintiffs, for the purpose of defrauding the public and taking from the plaintiffs the fair earnings of their skill, labor and enterprise. ); Chadwick v. Covell, 23 N.E. 1068, 1069 (Mass. 1890) (declaring that trademark law is intended to prevent one man from palming off his goods as another s, from getting another s business or injuring his reputation by unfair means, and perhaps, from defrauding the public ); Coats v. Holbrook, Nelson & Co., 7 N.Y. Ch. Ann. 586, 594 (N.Y. Ch. 1845) (stating that a party is not allowed to imitate the product of another and thereby attract to himself the patronage that without such deceptive use of such names... would have inured to the benefit of that other person ). 26 Where the defendant was using a technical trademark, courts were confident that the use had no legitimate explanation, so they were comfortable presuming that the defen-

10 2013] confusion isn t everything 261 As a result, early trademark and unfair competition law reached only a limited range of particularly egregious conduct acts by direct competitors (since only direct competitors could pass off goods as though they were the plaintiff s) that deceived consumers. And precisely because of this restricted scope, there were seldom cases where socially beneficial conduct ran the risk of liability. The modesty of the procompetitive norms embedded in traditional causes of action for trademark and unfair competition was generally sufficient to avoid conflict with broader competition or speech values. It should be no surprise, then, that there was little need for additional defenses in the traditional era, aside from trans-substantive equitable defenses like laches and acquiescence. 27 When courts ruled for defendants in cases that protected competition and speech values, they did so by enforcing the inherent limits of trademark rights. They might have invoked other values such as competition when rejecting a plaintiff s claim, 28 but they did so primarily to emphasize the reasons why trademark rights were so limited. Take, for example, Prestonettes, Inc. v. Coty, which many regard as an early application of the first sale doctrine. 29 In that case Prestonettes purchased genuine Coty toilet powder, subject[ed it] to pressure, add[ed] a binder to give it coherence and [sold] the compact in a metal case. 30 Prestonettes also rebottled genuine Coty perfume in smaller containers. 31 The district court had issued, and Prestonettes agreed to abide by, a decree that allowed Prestonettes to put upon the rebottled perfume Prestonettes, Inc., not connected with Coty, states that the contents are Coty s (giving the name of the article) independently rebottled in New York, every word to be in letters of the same size, color, type and general distinctiveness [and to] make compacts from the genuine loose powder of the plaintiff and to sell them with this label on the container: Prestonettes, Inc., not connected with Coty, states that the compact of face powder herein was independently compounded by dant intended to deceive. Where, on the other hand, a defendant s conduct was ambiguous where, for example the defendant might have been using a term to denote the class, grade, style, or quality of the articles courts required the plaintiff to prove the defendant intended to divert his trade. See, e.g., Diederich v. W. Schneider Wholesale Wine & Liquor Co., 195 F. 35, 37 (8th Cir. 1912); see also McKenna, supra note 5, at 1862 (describing the difference between trademark infringement and unfair competition in terms of proof of intent to deceive). 27 Even those doctrines were not complete defenses they applied to preclude damages from past infringement, but not to bar injunctions against future infringement. See, e.g., McLean v. Fleming, 96 U.S. 245, 251 (1877) ( Unreasonable delay in bringing a suit is always a serious objection to relief in equity; but cases arise in litigations of the kind before the court where the complainant may be entitled to an injunction to restrain the future use of a trade-mark, even when it becomes the duty of the court to deny the prayer of the bill of complaint for an account of past gains and profits. ). 28 See Kellogg Co. v. Nat l Biscuit Co., 305 U.S. 111, (1938); Singer Mfg. Co. v. June Mfg. Co., 163 U.S. 169, 185 (1896) U.S. 359 (1924). 30 Id. at Id. at 367.

11 262 notre dame law review [vol. 89:1 it from Coty s (giving the name) loose powder and its own binder. Loose powder-percent, Binder-per cent, every word to be in letters of the same size, color, type, and general distinctiveness. 32 But the Court of Appeals rejected that decree and issued an absolute preliminary injunction against the use of the [Coty] marks except on the original packages as marked and sold by [Coty], thinking that the defendant could not put upon the plaintiff the burden of keeping a constant watch. 33 The Supreme Court disagreed. To begin with, the Court thought it obvious that Prestonettes of course by virtue of its ownership had a right to compound or change what it bought, to divide either the original or the modified product, and to sell it so divided. The plaintiff could not prevent or complain of its stating the nature of the component parts and the source from which they were derived if it did not use the trade mark in doing so. 34 Thus, the only question was whether Prestonettes s use of Coty s trademark changed the result. And the answer to that question was clear once the Court focused on the limited nature of a trademark right: Then what new rights does the trade mark confer? It does not confer a right to prohibit the use of the word or words. It is not a copyright. The argument drawn from the language of the Trade Mark Act does not seem to us to need discussion. A trade mark only gives the right to prohibit the use of it so far as to protect the owner s good will against the sale of another s product as his.... When the mark is used in a way that does not deceive the public we see no such sanctity in the word as to prevent its being used to tell the truth. It is not taboo. 35 Prestonettes v. Coty did not carve out any special new first sale exemption from the scope of a trademark right. It did not allow the defendant to engage in conduct that caused the type of harm that was ordinarily actionable in order to vindicate some larger purpose of supporting competition. Instead the Court simply held that trademark law was not concerned with any use of a mark unless it deceived consumers into purchasing the defendant s product instead of the plaintiff s. Prestonettes s use of the Coty mark with the disclaimers described by the district court did not risk such passing off. Therefore, it was not infringing. If the name of Coty were allowed to be printed in different letters from the rest of the inscription dictated by the District Court a casual purchaser might look no further and might be deceived. But when it in no way stands out from the statements of facts that unquestionably the defendant has a right to communicate in some form, we see no reason why it should not be used collaterally, not to indicate the goods, but to say that the trade-marked product is a constituent in the article now offered as new and changed. As a 32 Id. 33 Id. 34 Id. at Id. (emphasis added) (citation omitted) (citing Canal Co. v. Clark, 80 U.S. 311, 327 (1872)).

12 2013] confusion isn t everything 263 general proposition there can be no doubt that the word might be so used. 36 Champion Spark Plug v. Sanders is similar, even though that case was decided in 1947, when trademark law had already begun to expand beyond its traditional limits. 37 In that case, the district court found that the defendants had infringed Champion s trademark and engaged in unfair competition when they sold used Champion spark plugs without making clear that those plugs were repaired and reconditioned. 38 The district court therefore enjoined the defendants from offering or selling any of [Champion s] plugs which had been repaired or reconditioned unless (a) the trade mark and type and style marks were removed, (b) the plugs were repainted with a durable grey, brown, orange, or green paint, (c) the word REPAIRED was stamped into the plug in letters of such size and depth as to retain enough white paint to display distinctly each letter of the word, (d) the cartons in which the plugs were packed carried a legend indicating that they contained used spark plugs originally made by [Champion] and repaired and made fit for use up to 10,000 miles by [defendants]. 39 The Second Circuit affirmed the findings of trademark infringement and unfair competition, but it modified the injunction: (a) it eliminated the provision requiring the trade mark and type and style marks to be removed from the repaired or reconditioned plugs; (b) it substituted for the requirement that the word REPAIRED be stamped into the plug, etc., a provision that the word REPAIRED or USED be stamped and baked on the plug by an electrical hot press in a contrasting color so as to be clearly and distinctly visible, the plug having been completely covered by permanent aluminum paint or other paint or lacquer; and (c) it elimi- 36 Id. at Importantly, the Court noted that the case was only one of trademark infringement and not unfair competition. The question therefore [was] not how far the court would go in aid of a plaintiff who showed ground for suspecting the defendant of making a dishonest use of his opportunities, but [was] whether the plaintiff ha[d] the naked right alleged to prohibit the defendant from making even a collateral reference to the plaintiff s mark. Id. at Champion Spark Plug Co. v. Sanders, 331 U.S. 125 (1947). 38 Id. at 126 ( The outside box or carton in which the plugs are packed has stamped on it the word Champion, together with the letter and figure denoting the particular style or type. They also have printed on them Perfect Process Spark Plugs Guaranteed Dependable and Perfect Process Renewed Spark Plugs. Each carton contains smaller boxes in which the plugs are individually packed. These inside boxes also carry legends indicating that the plug has been renewed. But respondent company s business name or address is not printed on the cartons. It supplies customers with petitioner s charts containing recommendations for the use of Champion plugs. On each individual plug is stamped in small letters, blue on black, the word Renewed, which at times is almost illegible. (footnote omitted)). 39 Id. at

13 264 notre dame law review [vol. 89:1 nated the provision specifying the precise legend to be printed on the cartons and substituted therefor a more general one. 40 On appeal to the Supreme Court, Champion argued that the Second Circuit had erred by allowing the defendant to keep the word Champion on the repaired or reconditioned spark plugs. 41 The Supreme Court rejected Champion s argument, implicitly holding that a party which adequately discloses the repaired or reconditioned status of authentic secondhand goods may use the original trademark to re-sell those goods. 42 Just as in Prestonettes, on which it relied extensively, the Court made clear that the defendant s use of the Champion mark with full disclosure of the goods reconditioned status simply fell outside the limited scope of Champion s trademark rights. 43 The Court put to one side the case of a manufacturer or distributor who markets new or used spark plugs of one make under the trade mark of another, 44 and it limited relief to disclosure of the defendant s repairing or reconditioning of the spark plugs because there had been no showing of fraud or palming off. 45 As this discussion illustrates, the outcomes in both Prestonettes and Champion depended entirely on the limited nature of trademark rights. In both cases, the Court found the defendant s conduct non-infringing because it was not the type of conduct that caused the specific harm trademark and unfair competition law addressed. 46 Neither decision relied on special excuses or exemptions or other limiting doctrines distinct from the prima facie case. And this approach was typical during that era outside of standard equitable 40 Id. at Id. at Id. at Id. 44 Id. at Id. at 130. There was fault, of course, in the defendant s original packaging, which the Court believed fell within the reach of trademark and unfair competition law because it did not adequately disclose that the spark plugs were used and were coming from a company other than Champion. Id. 46 We can see in Champion some evidence of the broadening of trademark law s reach that was already afoot. In the process of rejecting Champion s claim, the Court acknowledged that under Prestonettes, a second-hand dealer would get some advantage from the trade mark. Id. But, the Court said, such an advantage was wholly permissible so long as the manufacturer is not identified with the inferior qualities of the product resulting from wear and tear or the reconditioning by the dealer. Id. In Prestonettes the Court found that advantage permissible so long as it was not gained by deceiving consumers into believing that the used goods were the original goods. Prestonettes, Inc. v. Coty, 264 U.S. 359, 369 (1924). The Champion Court s slightly broadened view reflects courts increasing acceptance of a broader range of harms in trademark cases, and particularly their concern about the risk that the mark owner would be blamed for the low quality of goods bearing its mark. See Mark P. McKenna, Testing Modern Trademark Law s Theory of Harm, 95 IOWA L. REV. 63, (2009). We have more to say about this broadening, and its relation to limiting doctrines, below. See infra Section I.B.

14 2013] confusion isn t everything 265 doctrines, there really were no defenses in trademark law. 47 Certainly there were no analogues in these cases to modern doctrines like nominative fair use or any First Amendment defense. Nor were any other limiting doctrines needed. When it reached such a narrow range of conduct, there was little chance of trademark law conflicting seriously with other values. Each of our three fact patterns illustrates these inherent limits of earlier trademark rights. Competitors, resellers, and other third parties could promote their own goods and services using trademarks as long as they avoided passing off, as demonstrated in Prestonettes and Champion. And it would have been unimaginable in the early 1900s for a court to entertain a trademark claim against an artist who depicted a label in her painting or the producer of a motion picture that mentioned a brand name. Neither of those uses threatens trade diversion. As for product design, there was little need for a functionality doctrine because most courts categorically rejected unfair competition claims based on allegations that the defendant copied the design of a product, reasoning that unpatented designs were free for copying. 48 Even though those cases involved directly competing goods, courts could prevent trade diversion the only harm with which they were concerned by requiring clear labeling There were defenses in the formal sense because of the burden-shifting effect of federal registration. But these were not limiting doctrines that excused otherwise infringing behavior, distinct from the prima facie case. See William McGeveran, The Trademark Fair Use Reform Act, 90 B.U. L. REV. 2267, (2010). 48 See, e.g., Singer Mfg. Co. v. June Mfg. Co., 163 U.S. 169, 183, 185 (1896) (allowing the defendant to produce sewing machines with similar designs to the plaintiff s because the protecting patent had expired, even though consumers had come to associate the design with the plaintiff s company); Crescent Tool Co. v. Kilborn & Bishop Co., 247 F. 299, 301 (2d Cir. 1917) (expressly allowing the defendant to copy the plaintiff s [adjustable wrench] slavishly down to the minutest detail, as no one was entitled to monopolize any [unpatented] design or pattern, however trifling ); Flagg Mfg. Co. v. Holway, 59 N.E. 667, 667 (Mass. 1901) ( [T]he defendant ha[d] the same right that the plaintiff ha[d] to manufacture instruments in the present form, to imitate the arrangement of the plaintiff s strings or the shape of the body. In the absence of a patent the freedom of manufacture cannot be cut down under the name of preventing unfair competition. All that can be asked is that precautions shall be taken, so far as are consistent with the defendant s fundamental right to make and sell what it chooses, to prevent the deception which no doubt it desires to practice. (citations omitted)). 49 See, e.g., Kellogg Co. v. Nat l Biscuit Co., 305 U.S. 111, (1938) (noting that there was no evidence the defendant had attempted to pass off its product, as [t]he Kellogg cartons [were] distinctive. They d[id] not resemble those used by the plaintiff either in size, form, or color. And the difference in the labels [was] striking. The Kellogg cartons [bore] in bold script the names Kellogg s Whole Wheat Biscuit or Kellogg s Shredded Whole Wheat Biscuit so sized and spaced as to strike the eye as being a Kellogg product and also noting that, despite Kellogg s use of a picture of two shredded wheat biscuits in a bowl of milk which was quite similar to one of the plaintiff s registered trademarks... the name Kellogg was so prominent on all of the defendant s cartons as to minimize the possibility of confusion ); Singer, 163 U.S. at 204 (enjoining the defendant from using the word Singer or any equivalent thereto... without clearly and unmistakably stating in all said advertisements that the machines are made by the defendant, as distinguished from the

15 266 notre dame law review [vol. 89:1 While there were some outlying cases, 50 they were sufficiently rare that the functionality doctrine (which derived from the distinctions in these cases) only developed significantly later as trade dress gained greater trademark protection. 51 Early product design cases therefore reflect the general theme we have described: because trademark law conceived of its ends narrowly and reached a limited range of conduct, there was no real need for a special limiting doctrine to deal with mechanically functional features. Courts could simply observe the inherent limits of trademark rights to avoid conflict with the patent system and its competition norms. sewing machines made by the Singer Manufacturing Company ); Flagg Mfg., 59 N.E. at 667 ( [T]he plaintiff s right can be protected sufficiently by requiring the defendant s zithers to be clearly marked so as to indicate unmistakably that they are the defendant s and not the plaintiff s goods. ); see also Crescent Tool, 247 F. at 301 ( The defendant... may copy the plaintiff s goods slavishly down to the minutest detail; but he may not represent himself as the plaintiff in their sale. ). At times, Judge Hand, who authored the opinion in Crescent Tool, also seemed to suggest even the limited remedy of labeling would only be required with respect to nonfunctional features by which the article s source is distinguished, and where the public is moved in any degree to buy the [plaintiff s] article because of its source. Id. at See George G. Fox Co. v. Hathaway, 85 N.E. 417, 418 (Mass. 1908) (enjoining the defendants manufacturing and sale of bread loaves that mirrored the plaintiff s in size, shape, color, and general visual appearance despite the fact that the defendant sold its loaves under a different name and packaged its loaves in a broad paper band that was marked Hathaway s Log Cabin Bread. Finest Flavor, Malted in contrast to the plaintiff s small paper label containing the words Fox s Creamalt ); see also id. (rejecting the sufficiency of labeling because [t]here was nothing to show that the defendants business interests required the combination of this shape with the same size, color and general visual appearance as the plaintiff s loaves and noting that there were numberless shapes and sizes in which loaves of bread may be produced, and various peculiarities of appearance in color and condition of surface ). 51 Courts also often discussed refusals to register product design features in functionality terms. See, e.g., In re Oneida Cmty., 41 App. D.C. 260, , 1913 WL (D.C. Cir. 1913) (affirming the Trademark Office s refusal to register a circular or O-shaped film having distinct edges, on the back of the spoon bowl on the ground it was an attempt, under the guise of trademark registration, to obtain a monopoly of a functional feature of an article of manufacture when clearly no functional feature of a device is a proper subject for trademark registration (citation omitted)). Professor McCarthy suggests that these earlier decisions are precursors to modern functionality doctrine. See 1 MCCARTHY, supra note 21, 7:63. But these registration decisions do not use functionality in a modern sense they use that term merely to make clear that features of products themselves were not proper subject matter of trademark law: functional meant part of the product. See, e.g., Herz v. Loewenstein, 40 App. D.C. 277, 278, 1913 WL (D.C. Cir. 1913) (cancelling registration of a mark produced by a process of sealing the packages [that] produces a corrugated embossing on the ends of the wrappers, which constitutes the trademark on the ground that the mark [was] not placed upon the toothpicks, but [was] produced as the result of a distinct method of sealing the wrappers and as part of the manufactured wrappers, [was] not subject to registration as an arbitrary mark ). Moreover, those cases only dealt with registrability; they did not deal with protection via unfair competition claims for unregistered trade dress.

16 2013] confusion isn t everything 267 This model only worked when trademark rights remained narrow. But as the next Section explains, the subsequent massive expansion of trademark rights disrupted courts ability to rely on inherent limits in the prima facie case. B. The Emergence of the Confusion Rationale Beginning around 1920, courts began to see traditional trademark law s focus on trade diversion as overly constrained. The increasing complexity of manufacturing and marketing practices exerted increasing pressure throughout the twentieth century to broaden the scope of trademark rights. Courts responded to this pressure by recognizing a much wider range of uses of a mark as potentially infringing and accepting a wider range of designations as protectable subject matter. First, courts began to find infringement when use of a mark caused consumers to think either (1) that the plaintiff actually produced the defendant s goods or (2) that the plaintiff sponsored the defendant s goods or was affiliated with their producer. In Vogue Co. v. Thompson-Hudson Co., for example, the court held that the defendant s use of The Vogue Hat Company to sell hats infringed the plaintiff s rights in the VOGUE mark for magazines. 52 Even if consumers would not believe the hat company was the magazine publisher, it was enough if they might incorrectly believe there was a relationship between the two. 53 We mean to emphasize two aspects of this expansion. First, and most obviously, courts extended trademark infringement beyond the context of directly competing goods, deemphasizing trade diversion as unfair competition s defining harm. Second, courts elevated consumer confusion, making it, rather than deception, the central inquiry. 54 Together F. 509, 512 (6th Cir. 1924). 53 Id. 54 In the late nineteenth and early twentieth centuries, courts regularly denied relief in cases in which they acknowledged that consumers were likely to be confused. In Canal Co. v. Clark, for example, the Supreme Court refused to prevent the defendant from truthfully describing his coal as having originated from Lackawanna despite the plaintiff s prior use of that geographically descriptive term. 80 U.S. 311, 324 (1872). It did so even though it recognized there was a risk of consumer confusion, because confusion was not enough. Id. at 327 ( [I]f it is just as true in its application to his goods as it is to those of another who first applied it, and who therefore claims an exclusive right to use it, there is no legal or moral wrong done. Purchasers may be mistaken, but they are not deceived by false representations, and equity will not enjoin against telling the truth. ). American Washboard Co. v. Saginaw Manufacturing Co., 103 F. 281 (6th Cir. 1900), is also illustrative here. In that case, the plaintiff manufactured washboards with aluminum coated facings and sold the washboards under the aluminum trade name. Id. at 282. The defendant also designated its washboards as aluminum, even though it actually made its products out of zinc. Id. at 283. While there was little dispute that the defendant had in fact misrepresented the nature of its goods, the court denied injunctive relief. Id. at 281. Since aluminum was merely the descriptive title of a kind of washboard, no single producer could claim the term as its own. Id. at 282. Without additional acts beyond use of the aluminum trade name, the plaintiff could not claim that the defendant was palming off its goods as those